RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

“THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING PRACTICES”: USING THE ONE- PERSON, ONE-VOTE PRINCIPLE TO COMBAT PARTISAN GERRYMANDERING

∗ ALLISON J. RIGGS ANITA S. EARLS

INTRODUCTION In 2013 and 2015, the North Carolina General Assembly, controlled by a Republican supermajority, passed local bills restructuring Wake County’s Board of Education and Board of County Commissioners, respectively. These bills were passed over the loud objection of the majority of the Wake County delegation and, indeed, the vast majority of the county’s voters. Wake County, home to the capital city of Raleigh and part of the state’s Research Triangle Park, has long been a progressive bastion in the state, with Democratic registration significantly outnumbering Republican registration in the county, and Democratic candidates regularly outperforming Republican candidates. With carefully manipulated district lines, those local bills would have ensured Republican control of both boards, despite the strong Democratic leaning in the county. Critically, in such a heavily Democratic county, the only way that the General Assembly could achieve such a drastic partisan skew was to overpopulate the Democratic-leaning districts and underpopulate the Republican-leaning districts, right up to what the legislature treated as a ten percent total deviation safe harbor. But the Supreme Court’s

Copyright © 2017 Allison J. Riggs and Anita S. Earls. ∗ Anita Earls is the Executive Director of the Southern Coalition for Social Justice in Durham, North Carolina. Allison Riggs is a senior attorney at the Southern Coalition for Social Justice, specializing in voting rights. The authors of this article served as counsel for the organizational and individual plaintiffs in the Wake County cases discussed in this article. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

24 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3

“One-Person, One-Vote” (“OPOV”) jurisprudence makes clear that the equal weighting of citizens’ votes is vital to the functioning of our democracy. As Justice Stevens noted in his concurrence to the Supreme Court’s summary affirmance in Cox v. Larios,1 “the equal-population principle remains the only clear limitation on improper districting practices, and we must be careful not to dilute its strength.” 2 In the fifty or so years in which federal courts have been willing to meaningfully review legislatively-enacted redistricting plans, two race- neutral approaches have emerged to define and explain why plans might be unconstitutionally unfair. The first is the equal population approach under the Equal Protection Clause,3 and the second is the partisan-gerrymandered case, which belief prohibits districting systems that are rigged to ensure that one political party remains in power under the First and Fourteenth Amendments.4 One of those—the equal population approach—has resulted in more success and a much more thorough jurisprudence. Judges have latched onto the idea that every voter should be able to cast a vote that is weighted the same as every other in a jurisdiction. As scholars have noted, though, “[o]ne person, one vote’s individualistic rhetoric may have come to obscure its original purposes of combating entrenchment and safeguarding majority rule.”5 But these two approaches are not unrelated. One recent case brought as an equal population challenge could significantly develop the jurisprudence of partisan gerrymandering cases. In a country marked by increasing political polarization, such overstepping as seen in the Wake County, North Carolina, case is likely to be repeated across the country. Voting rights litigants achieved an important victory in 2016 when the Fourth Circuit ruled in consolidated challenges to those two local bills restructuring those county boards— Wright v. North Carolina6 and Raleigh-Wake Citizens Association v. Wake County Board of Elections7—that the new redistricting plans violated the Fourteenth Amendment’s equal population guarantee. This Article examines the interplay between OPOV litigation and

1. Cox v. Larios, 542 U.S. 947 (2004). 2. Id. at 949–50. 3. Reynolds v. Sims, 377 U.S. 533, 577 (1964). 975 F. Supp. 2d 539 (E.D.N.C. 2014). 4. Vieth v. Jubelirer, 541 U.S. 267, 314–15 (2004) (Kennedy, J., concurring); Davis v. Bandemer, 478 U.S. 109, 143 (1986). 5. Samuel Issacharoff & Pamela S. Karlan, Where to Draw the Line?: Judicial Review of Political Gerrymanders, 153 U. PA. L. REV. 541, 544 n.17 (2004). 6. 975 F. Supp. 2d 539 (E.D.N.C. 2014). 7. 827 F.3d 333 (4th Cir. 2016). RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 25 partisan gerrymandering cases using the Wake County case as the vehicle for understanding the relationship between the two legal approaches. The Article examines the genesis of the Wake County challenges, focusing on how the litigants successfully gathered evidence to demonstrate that illegitimate partisan considerations drove the population deviations. The Article concludes by positing how, even as partisan gerrymandering cases seem to finally be bearing fruit, the OPOV principle still provides the single most important “limitation on improper districting practices” and litigation under that theory should be pursued. This approach will create a legal atmosphere where partisan gerrymandering claims are more likely to succeed.

I. BACKGROUND JURISPRUDENCE Before embarking on this case study, it is important to trace the jurisprudential development of two legal theories significant to this case: the one-person, one-vote guarantee and unconstitutional partisan gerrymandering.

A. Development of the One Person, One Vote Principle The involvement of the federal judiciary in ensuring that electoral districts are evenly populated, and thus all voters across a jurisdiction cast an evenly-weighted vote, is relatively recent. When it came to apportionment and redistricting, the Supreme Court had long been wary of wading into difficult questions relating to political representation, and Justice Felix Frankfurter once famously cautioned, “[c]ourts ought not to enter this political thicket.”8 That all changed, though, in 1962, when the Court authorized federal courts to begin entering the “political thicket” in Baker v. Carr.9 In Baker, voters in Tennessee challenged the state’s 1901 law that apportioned the members of the General Assembly among the state’s ninety-five counties, where each county was apportioned at least one representative and one senator.10 Plaintiffs alleged that the uneven divvying created a “debasement of their votes” in violation of the Equal Protection Clause of the Fourteenth Amendment.11 Because courts had not previously required jurisdictions to adjust district lines after a federal census or other population enumeration, by the 1960s,

8. Colegrove v. Green, 328 U.S. 549, 556 (1946). 9. 369 U.S. 186 (1962). 10. Id. at 188–89. 11. Id. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

26 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3 many state legislatures were starkly malapportioned, and Tennessee was no exception. In his concurrence in Baker, Justice Douglas noted that “a single vote in Moore County, Tennessee, is worth [nineteen] votes in Hamilton County, that one vote in Stewart or in Chester County is worth nearly eight times a single vote in Shelby or Knox County.”12 The Supreme Court in Baker held that the political question doctrine does not foreclose review of redistricting cases simply because political rights are affected.13 With courts now authorized to review constitutional challenges to reapportionment plans, the Supreme Court over the next two years articulated when such plans could run afoul of the Constitution, leading to the creation of the OPOV rule. In Gray v. Sanders,14 the Court noted: How then can one person be given twice or ten times the voting power of another person in a statewide election merely because he lives in a rural area or because he lives in the smallest rural county? Once the geographical unit for which a representative is to be chosen is designated, all who participate in the election are to have an equal vote—whatever their race, whatever their sex, wherever their occupation, whatever their income, and whatever their home may be in that geographic unit. This is required by the Equal Protection Clause of the Fourteenth Amendment.15 It boiled down to this: “[t]he conception of political equality . . . can only mean one thing—one person, one vote.”16 The next year, in Reynolds v. Sims,17 striking down Alabama state legislative districts as unconstitutionally malapportioned, the Court explained that “[l]egislators represent people, not trees or acres,”18 so Alabama’s desire to assign one senator to each county and to ensure that every county had at least one state representative was not justifiable, and instead only enshrined geographical and regional favoritism.19 The Court further fleshed out the harm such efforts cause:

12. Id. at 245 (Douglas, J., concurring). 13. Id. at 209. 14. 372 U.S. 368 (1963). 15. Id. at 379. 16. Id. at 381. 17. 377 U.S. 533 (1964). 18. Id. at 562. 19. Id. at 543–44, 563. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 27

If a [s]tate should provide that the votes of citizens in one part of the [s]tate should be given two times, or five times, or ten times the weight of votes of citizens in another part of the [s]tate, it could hardly be contended that the right to vote of those residing in the disfavored areas had not been effectively diluted.20 Thus, the Court reaffirmed the constitutional harm that can follow if anything other than population is used as the basis of apportionment of electoral districts. Avery v. Midland County21 then extended this rule to local redistricting.22 But how can a jurisdiction redistrict to achieve this desired OPOV result? The Court has emphasized that while “[m]athematical exactness” in population amongst electoral districts is not a workable standard for state and local government redistricting, governments must nonetheless “make an honest and good faith effort to construct districts . . . as nearly of equal population as is practicable.”23 The rule that has emerged is that where redistricting plans have a maximum or overall population deviation over ten percent, that deviation will create a prima facie case of discrimination and thus must be justified by the state.24 Under Roman v. Sincock,25 plans that have overall deviations under ten percent will not be immune from attack, but a plaintiff cannot rely on the deviation percentage alone to establish a constitutional harm and must provide further evidence to make his case.26 Importantly, the Court then explained that “minor deviations” in redistricting plans are only acceptable insofar as they are “free from any taint of arbitrariness or discrimination.”27 Relevant here, the Fourth Circuit gave that Roman rule some depth, explaining that in order to survive summary judgment plaintiffs must produce “credible evidence to establish that the apportionment plan at issue . . . was the product of bad faith, arbitrariness, or invidious discrimination.”28

20. Id. at 562. 21. 390 U.S. 474 (1968). 22. Id. at 485. 23. Reynolds, 377 U.S. at 577. 24. Brown v. Thompson, 462 U.S. 835, 842–43 (1983). 25. 377 U.S. 695 (1964). 26. See id. at 710 (noting the equal population “problem does not lend itself to any such uniform formula, and it is neither practicable nor desirable to establish rigid mathematical standards for evaluating the constitutional validity” of allegedly malapportioned plans). 27. Id. 28. Daly v. Hunt, 93 F.3d 1212, 1222 (4th Cir. 1996). RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

28 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3

B. Development (or a Lack) of a Justiciable Standard for Partisan Gerrymandering Claims Davis v. Bandemer29 was one of the first challenges to a reapportionment plan that arrived at the Supreme Court solely on the claims that it was an unconstitutional partisan gerrymander. The challengers in Bandemer alleged that the state legislative districts for the Indiana legislature had the intent and effect of securing Republican control of the legislature, to the detriment of Democratic voters, even though the state was considered a swing state and, indeed, Democratic candidates had receive majority of votes cast in both chambers in the first election held under the challenged plan.30 A three-judge panel in the district court found that the redistricting plans did, in fact, constitute unconstitutional partisan gerrymanders.31 The Court’s consideration of the case produced a splintered set of opinions, but six justices held that political gerrymandering claims are justiciable.32 Beyond that, there was little agreement other than the fact that the district court in Bandemer got it wrong, largely because the Court concluded that because the challenged plan did not produce proportional representation for the political parties, challengers had failed their burden in establishing discriminatory intent.33 A plurality did agree that a partisan gerrymander violates the Equal Protection Clause only on a showing of “both intentional discrimination against an identifiable political group and an actual discriminatory effect on that group.”34 Justice Powell proposed a totality of circumstances approach, arguing that courts should look to a number of factors in analyzing partisan gerrymander, including “the legislative procedures by which the challenged redistricting was accomplished and the intent behind the redistricting; the shapes of the districts and their conformity with political subdivision boundaries; and ‘evidence concerning population disparities and statistics tending to show vote dilution.’”35 The Court declined this standard, though, worrying that it would “invite[] judicial interference in legislative districting whenever a political party suffers at the polls.”36

29. 478 U.S. 109 (1986). 30. Id. at 115. 31. Bandemer v. Davis, 603 F. Supp. 1479 (S.D. Ind. 1984). 32. See Davis, 478 U.S. at 143. 33. See id. at 129–31. 34. Id. at 127. 35. Id. at 138 (quoting id. at 173 (Powell, J., concurring in part and dissenting in part)). 36. Id. at 142. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 29

Significantly, though, in concluding that partisan gerrymandering is justiciable, the Court spent a substantial amount of time connecting the question before its OPOV jurisprudence. The Court noted that, “in formulating the one person, one vote formula, [it] characterized the question posed by election districts of different sizes as an issue of fair representation,” and that if fair representation was central to the Reynolds Court’s finding that OPOV cases were justiciable, the same conclusion is warranted in the political gerrymandering realm.37 Thus, while the Court in Bandemer did not provide any meaningful relief to the specific litigants before it, it left the door open for future attack on partisan gerrymandering and linked the ultimate inquiry—fairness—to the Court’s concerns in the OPOV context. After Bandemer, litigants struggled to develop a manageable standard to offer to the Court for analyzing a partisan gerrymandering case, and defendants continued to appeal to the Court’s continued division on whether such cases were justiciable at all. These tensions again came to a head again in Vieth v. Jubelirer.38 In reviewing a partisan gerrymandering challenge to Pennsylvania’s congressional districts, the Court again was badly splintered. A plurality of the court, in voting to dismiss the case, agreed that partisan gerrymandering was non- justiciable and thus would have overruled Bandemer,39 but Justice Kennedy declined to join the plurarity opinion, instead writing separately to concur and affirm the district court’s dismissal of the case.40 Instead, he continued to assert that political gerrymandering claims were justiciable notwithstanding the Court’s inability to decide on a proper standard, yet, for judging those claims.41 Like the plurality and the court below, Justice Kennedy rejected “the fairness principle appellants propose . . . that a majority of voters in the Commonwealth should be able to elect a majority of the Commonwealth’s congressional delegation.”42 Instead, in pointing future litigants in a direction that would someday produce a manageable standard, he noted that “[a] determination that a gerrymander violates the law must rest . . . on a conclusion that the classifications, though generally permissible, were applied in an invidious manner or in a manner

37. Id. at 123–24. 38. 541 U.S. 267 (2004). 39. Id. at 281. 40. Id. at 306 (Kennedy, J., concurring). 41. Id. at 307. 42. Id. at 308. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

30 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3 unrelated to any legitimate legislative objective.”43 Finally, Justice Kennedy again reiterated the significance of the equal population jurisprudence to his belief that partisan gerrymandering was justiciable.44 Further analogizing the OPOV jurisprudence, Kennedy noted that the Court in Baker recognized the justiciability of malapportionment cases before the Court actually articulated a specific standard in Reynolds.45 Thus, opponents of partisan gerrymandering, while making no progress, lived to fight another day. The Court’s most recent pronouncement on partisan gerrymandering came from a complex case out of Texas following the state’s mid-decade redistricting—League of United Latin Am. Citizens v. Perry (“LULAC”).46 There, a federal court had redrawn the state’s congressional districts after the 2000 census, when the legislature, under split-partisan control could not agree on a plan.47 However, Republicans took control of the legislature in 2002, and in 2003, redrew the state’s congressional districts to give Republicans a strong majority in the delegation.48 Because population deviations between congressional districts are generally required to be as close to zero as possible,49 the mid-decade redistricting plan did not create any opportunity to challenge the deviations on OPOV grounds. That did not, however, keep creative litigants from arguing that the redistricting process still violated the equal population guarantee.50 The Court this time did not even bother revisiting the justiciability question, but did “examine whether appellants’ claims offer the Court a manageable, reliable measure of fairness for determining whether a partisan gerrymander violates the Constitution.”51 It concluded they did not.52 Appellants in LULAC offered two distinct theories of partisan gerrymandering. The first was a “sole motivation” standard, arguing

43. Id. at 307. 44. See id. at 310 (“Our willingness to enter the political thicket of the apportionment process with respect to one-person, one-vote claims makes it particularly difficult to justify a categorical refusal to entertain claims against this other type of gerrymandering.”). 45. Id. 46. 548 U.S. 399 (2006). 47. Id. at 411 (2006). 48. Id. at 412. 49. See Wesberry v. Sanders, 376 U.S. 1, 7–8 (1964). 50. See LULAC, 548 U.S. at 413 (presenting the court with two different theories on why mid-decade redistricting with decennial census data was problematic). 51. Id. at 414. 52. Id. at 417. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 31 that engaging in “mid-decennial redistricting, when solely motivated by partisan objectives, violates equal protection and the First Amendment because it serves no legitimate public purpose and burdens one group because of its political opinions and affiliation.”53 Appellants focused on motivation, particularly with mid- decade redistricting, because it would obviate the need to deal with the messy proposition of measuring the discriminatory effects of partisan gerrymandering.54 Ultimately though, the Court concluded that while the legislature decided to redistrict “with the sole purpose of achieving a Republican congressional majority,” such partisan goals did not “guide every line it drew,” because some districts had lines informed by “mundane and local interests.”55 More importantly here, the second standard focused on OPOV grounds. Specifically, appellants argued that because the census data would be further outdated and inaccurate mid-decade than it was when first used for decennial redistricting, the decision to redistrict mid- decade was a strong indicator that bad faith motivated the process.56 Therefore, the practice of mid-decade redistricting meant that the legislature had not made a good faith effort to achieve equal population.57 Unfortunately, the Court concluded that “[t]his is a test that turns not on whether a redistricting furthers equal-population principles but rather on the justification for redrawing a plan in the first place. In that respect appellants’ approach merely restates the question whether it was permissible for the Texas Legislature to redraw the districting map.”58 Thus, as of 2006, attempts to gain any traction in the Supreme Court in challenging partisan gerrymandering remain largely unsuccessful, but such claims are still not categorically barred.

C. The Larios Case – A Successful Challenge to a Plan with Less than 10% Deviation Because the burden on challengers of plans with less than ten percent overall deviation requires them to produce additional evidence besides the deviation to prove a plan is unconstitutional, challenges to plans with deviations under ten percent have been relatively rare. However, a case from Georgia in the early 2000s helped to define the

53. Id. at 416–17. 54. Id. at 417. 55. Id. at 417–18. 56. Id. at 420–21. 57. Id. at 421. 58. Id. at 422 (internal citations omitted). RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

32 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3 types of evidence that challenges can rely upon to invalidate a plan with a total deviation under ten percent. In Larios v. Cox,59 Republican voters challenged on OPOV grounds Georgia’s 2001 and 2002 redistricting plans for the State House and Senate, respectively.60 Democrats controlled the legislature at the time, but the 2010 census revealed that areas dominated by Republican voters, including suburban areas surrounding Atlanta, grew at a much larger rate than southern Georgia, which was rural and more Democratic.61 The enacted House and Senate plans each had an overall deviation of 9.98%.62 In both, the underpopulated districts were primarily Democratic-leaning, and the overpopulated districts were primarily Republican-leaning.63 Also in both, most of the underpopulated districts were either in southern Georgia or within the urban parts of Atlanta—the more Democratic-leaning regions of the state.64 In the House plan, fifty percent of all Republican incumbents were drawn into districts with another incumbent, but only nine percent of Democratic incumbents.65 In the Senate, the plan paired forty-two percent of all Republican incumbents, but only six percent of Democratic incumbents.66 A three-judge panel at the district court level found that both direct and circumstantial evidence left “no doubt that a deliberate and systematic policy of favoring rural and inner-city interests at the expense of suburban areas north, east and west of Atlanta led to a substantial portion of the 9.98% population deviations in both of the plans.”67 It found that this systematic approach of over- and under- populating districts based on the voting patterns in the districts “led to a significant overall partisan advantage for Democrats in the electoral maps.”68 The court thus concluded that “[t]he population deviations in the Georgia House and Senate Plans are not the result of an effort to further any legitimate, consistently applied state policy.”69 Furthermore, the reasons for the deviations—to allow rural and urban

59. 300 F. Supp. 2d 1320 (N.D. Ga. 2004). 60. Id. at 1324–25. 61. Id. at 1322–23. 62. Id. at 1326–27. 63. Id. 64. Id. 65. Id. at 1326. 66. Id. at 1327. 67. Id. 68. Id. at 1331. 69. Id. at 1338. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 33 regions of the state to hold onto legislative influence beyond what their populations should allow and to protect Democratic incumbents while undermining Republican incumbents—did not withstand Fourteenth Amendment scrutiny.70 On direct appeal, the Supreme Court summarily affirmed the ruling.71 Of course, a summary affirmance has limited precedential value, so “should not be understood as breaking new ground but as applying principles established by prior decisions to the particular facts involved.”72 However, the summary affirmance was still significant, as explicated by the concurrences filed with the order. Justice Stevens, with whom Justice Breyer joined in concurring, wrote that the district court’s findings disclosed two reasons for the unconstitutional population deviations in the enacted plans—the favoring of rural and urban over suburban interests, and an “intentional effort to allow incumbent Democrats to maintain or increase their delegation, primarily by systematically underpopulating the districts held by incumbent Democrats, by overpopulating those of Republicans, and by deliberately pairing numerous Republican incumbents against one another.”73 Even more significantly, Justice Stevens explained: In challenging the District Court’s judgment, appellant invites us to weaken the one-person, one-vote standard by creating a safe harbor for population deviations of less than 10 percent, within which districting decisions could be made for any reason whatsoever. The Court properly rejects that invitation. After [Vieth v. Jubelirer], the equal-population principle remains the only clear limitation on improper district practices, and we must be careful not to dilute its strength.74 Ironically, the concurrence noted that had the Court adopted a standard for partisan gerrymandering in Vieth, what happened in Georgia would have surely have satisfied any proposed standard.75 Significantly, at least in Justice Stevens and Justice Breyer’s minds, OPOV claims and partisan gerrymandering claims are not unrelated— both can present opportunities to restrain the toxic effect of rank partisanship in redistricting. Even where the partisan gerrymandering jurisprudence has not yet developed to the point where plans can be

70. Id. 71. Cox v. Larios, 542 U.S. 947 (2004). 72. Mandel v. Bradley, 432 U.S. 173, 176 (1977). 73. Cox, 542 U.S. at 947 (Stevens, J., concurring) (quoting Larios, 300 F. Supp. 2d at 1329). 74. Id. at 949–50 (internal citations omitted). 75. Id. at 950. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

34 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3 invalidated under that theory, the well-established OPOV rule can fill the gap. Larios thus created instructive guidance for the case examined in this Article. Of course, Larios was not the first or only case invalidating a redistricting plan with less than ten percent overall deviation. In Hulme v. Madison County,76 a district court found that a districting scheme for a county board of commissioners, with a population deviation of just 9.3%, violated the Equal Protection Clause.77 The court found the deviations were arbitrary and discriminatory because the main cause of the deviations was the intent to “create districts that would not simply disadvantage Republican members of the Board, but ‘cannibalize’ their districts to the greatest extent possible.”78 The chairman of the county board provided evidence of such discriminatory intent when he infamously explained to an opponent of the plan: “We are going to shove it [the map] up your f ass and you are going to like it, and I’ll f any Republican I can.”79 Few litigants enjoy such frank admissions from proponents of challenged redistricting plans. After Larios, a three-judge panel in Texas found a discriminatory pattern in challenged districts with population deviations falling below even five percent. The court in Perez v. Perry80 applied a preliminary injunction standard as the basis for drawing an interim, court-ordered plan for the State House where the enacted plan employed a scheme of overpopulating and underpopulating districts based on the partisan preferences of the voters in the district.81 The federal court noted that: [E]nacted HD 41—the only district in Hidalgo County that has a realistic chance of electing a Republican—is substantially underpopulated (by 4.41% from the ideal district size), but the rest of the districts in the county are substantially overpopulated (2.83%, on average). This apparently systematic overpopulation of Democrat districts and underpopulation of the one possible Republican district presents serious concerns under Larios v. Cox . . . . Thus, the plaintiffs have a likelihood of success on the merits on their one-person, one-vote claims with respect to HD 41.82

76. 188 F. Supp. 2d 1041 (S.D. Ill. 2001). 77. Id. at 1051. 78. Id. 79. Id. 80. No. 5:11-cv-360, 2012 U.S. Dist. LEXIS 190609 (W.D. Tex. Mar. 19, 2012). 81. Id. at *53–*54. 82. Id. at *55–*56 (internal citations omitted). RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 35

Thus, even with what might arguably described as minimal population deviations, the Texas federal court invalidated districts for being constitutionally malapportioned.

III. BACKGROUND TO THE WRIGHT/RWCA CASES In 2010, Republicans took control of the North Carolina General Assembly for the first time in over one hundred years.83 When they redistricted the state legislative seats the next year, they created a supermajority for themselves that has persisted through the decade.84 For decades before Republicans took control of the legislature, the standard practice for using local bills to pass legislation relevant only to a small number of counties in the state was that such bills were usually only pursued at the request of local elected bodies and with the unanimous support of the local legislative delegation.85 Thus, if City X wanted to restructure its municipal elections, it might be faster, if the request were noncontroversial, to ask the legislature to enact a local bill to achieve that end result more quickly. While of course there are always exceptions to the rule, this represented the long-standing practice of the legislature. After the Republican legislature had secured its own future political success, though, it turned to tinkering with how members of county government were elected, and one of the first of such efforts was directed at the Wake County Board of Elections.

A. Wake County Politics and the 2013 Legislative Process Wake County is home to North Carolina’s capital city, Raleigh, and hosts, with neighboring counties, North Carolina’s Research Triangle Park, a strong attractor of business and talent in the region. The county has grown tremendously over the last decade. Indeed, concomitant with that growth, the student population in Wake County Public School System grew 46.68% from 2000 to 2010, from 98,772 to 143,289

83. Kim Severson, G.O.P.’s Full Control in Long-Moderate North Carolina May Leave Lasting Stamp, N.Y. TIMES (Dec. 11, 2012), http://www.nytimes.com/2012/12/12/us/politics/gop- to-take-control-in-long-moderate-north-carolina.html. 84. Dan Boylan, Republicans Headed Toward Keeping Veto-proof Majorities, NEWS & OBSERVER (Nov. 8, 2016), http://www.newsobserver.com/news/politics-government/election/ article113483393.html. 85. See, e.g., UNC CHAPEL HILL SCHOOL OF GOV’T, THE NORTH CAROLINA GENERAL ASSEMBLY: OVERVIEW OF THE STRUCTURE, LEGISLATION AND PROCESS OF NORTH CAROLINA’S LEGISLATURE 9 https://www.sog.unc.edu/sites/www.sog.unc.edu/files/cours e_materials/General %20Assembly%20Handout%20-%20Master_0.pdf (describing usual method of enacting local bills). RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

36 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3 students.86 This growth resulted in fairly significant school overcrowding, and this problem was addressed by introducing year- round schools, increasing the county’s magnet school program, and continuing the county’s socioeconomic diversity busing program.87 The latter, in particular, caused policy debate in the county and on the nine- member school board. Prior to 2013, the nine members of the school board were elected, via non-partisan elections, from single member districts. For much of the 2000s, progressives controlled the non-partisan board.88 The decision to adopt a socioeconomic diversity assignment policy in 2000 was supported by both Democrat and Republican school board members. But in 2009, partisan debates over the merits of the county’s socioeconomic diversity policy came to a head when a slate of Republican candidates overwhelmingly swept four of the nine school board seats up for election, and with the seats already held by Republicans, took control of the Board.89 The new Republican majority took immediate steps to replace the socioeconomic diversity assignment plan with a neighborhood student assignment policy. Huge public outcry ensued. The State Conference of the NAACP and parents of Wake County students filed a complaint under Title VI of the Civil Rights Act, arguing that neighborhood schools would result in highly segregated schools.90 Community organizers began work on ensuring that this change in approach would be central in voters’ mind when they voted in the 2011 school board elections.91 And in that election, registered Democrats won four of the five seats on the ballot in 2011, with an unaffiliated but progressive candidate winning the fifth seat.92 The results of this election were seen as a definitive rejection of conservative education policies and neighborhood school assignment plans.93

86. See Raleigh Wake Citizens Ass’n v. Wake Cnty. Bd. of Elections (RCWA I), 166 F. Supp. 3d 553, 564 (E.D.N.C. 2016). 87. Transcript of Testimony of School Board Member Bill Fletcher, Dec. 16, 2015, at 85:18- 86:23 (all trial transcripts on file with author). 88. RWCA I, 166 F. Supp. 3d at 564. 89. Karen McMahan, Democrats Mobilize to Win Back Majority on Wake School Board, CAROLINA J. (July 28, 2011), https://www.carolinajournal.com/news-article/democrats-mobilize- to-win-back-majority-on-wake-school-board/. 90. Sloane Heffernan, NAACP files complaint against Wake County schools, WRAL.COM (Sept. 25, 2010) http://www.wral.com/news/local/story/8348916/. 91. Transcript of Testimony of Rev. Earl Johnson, Dec. 16, 2015, at 24:17-25:21, RWCA I, 166 F. Supp. 3d 553 (No. 5:15-CV-156-D). 92. RWCA I, 166 F. Supp. 3d at 568. 93. Transcript of Testimony of School Board Member Christine Kushner, Dec. 16, 2015, at RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 37

Significantly, in that two-year period where Republicans controlled the school board, the board was tasked with redrawing its electoral districts to even out the population differences between the districts highlighted by the 2010 census results.94 The board hired a law firm frequently hired by Republican groups, and the firm drew a plan that would undoubtedly favor Republicans,95 but maintained the current method of election, drew compact districts, and achieved an overall population deviation of just 1.66%.96 Parents and community organizations were upset with this ploy, but of course, as it turned out, the subtle favoritism employed was not enough to overcome community outrage with the changes to the student assignment policy, and Republicans overwhelmingly lost in the 2011 elections.97 After the 2011 election results, the Wake County Board of County Commissioners, responsible for allocating funds to the school board and temporarily controlled by a narrow Republican majority, set as one of its legislative goals seeking local legislation to ensure that a majority of the School Board was elected on an at-large basis.98 Senate Bill 325 was filed in the North Carolina General Assembly on March 13, 2013, by Republican Senator Neal Hunt of Wake County.99 The billed called for redistricting the school board using seven numbered single-member districts and two lettered single- member “super districts” that overlap the seven numbered districts, splitting Wake County into a doughnut shape with most of Raleigh in the center.100 Those super districts are displayed below:

104:2-7, RWCA I, 166 F. Supp. 3d 553 (No. 5:15-CV-156-D). 94. RWCA I, 166 F. Supp. 3d at 566. 95. Transcript of Testimony of Sen. Dan Blue, Dec. 16, 2015, at 51:14-24, RWCA I, 166 F. Supp. 3d 553 (No. 5:15-CV-156-D). 96. RWCA I, 166 F. Supp. 3d at 567. 97. Id. at 568. 98. Transcript of Testimony of Comm’r James West, Dec. 17, 2015, at 26:3-27:21, RWCA I, 166 F. Supp. 3d 553 (No. 5:15-CV-156-D). 99. RWCA I, 166 F. Supp. 3d at 568. 100. Id. at 569. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

38 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3

RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 39

The bill created a system in which each Wake County voter would vote for two school board representatives: one from the numbered district in which the voter resided, and one from the lettered district in which the voter resided.101 The plan split thirty-one precincts, whereas the 2011 plan passed by the Republican school board split only ten precincts.102 Significantly, the super districts had an overall population deviation of 9.8%.103 The outer “doughnut” district, comprised primarily of the more conservative and suburban parts of the county, was underpopulated by 4.9% and the inner urban district was overpopulated by 4.9%.104 Additionally, among the seven single- member districts, the more urban districts primarily within Raleigh were overpopulated whereas the suburban districts were again underpopulated.105 Moreover, numerous incumbents who favored the county’s socioeconomic diversity plan were paired in the proposed districts, or drawn into heavily Republican-leaning districts. Reconstituted election results demonstrated that Republican candidates would win in five of the nine new districts, thus swinging control from Democrats to Republicans. The bill passed the Senate in April and the House in June, and was ratified on June 13, 2013.106 As expected, proponents of Senate Bill 325 were not frank about the reasons motivating the unwanted legislation.107 Proponents alleged three main goals of the legislation: to give suburban voters a larger voice in school board governance;108 to allow Wake County voters to elect two members instead of just one;109 and to better align electoral districts with student assignment districts. These justifications would become central part of later litigation.110 Interestingly, though, Republican supporters of the bill rejected amendments from Democratic legislators that would have made the super districts true

101. Id. at 568. 102. Id. at 605. 103. Id. at 559. 104. Id. at 571. 105. Id.; see also Transcript of Testimony of Anthony Fairfax, Dec. 16, 2015, at 138:6-21, RWCA I, 166 F. Supp. 3d 553 (No. 5-15-CV-156-D). 106. RWCA I, 166 F. Supp. at 573. 107. Raleigh Wake Citizens Ass’n v. Wake Cnty. Bd. of Elections (RCWA II), 827 F.3d 333, 349 (4th Cir. 2016) (noting “the stated reasons for the redistricting as pretextual”). 108. Transcript of Testimony of Sen. Josh Stein, Dec. 17, 2015, at 20:25-21:4, RWCA I, 166 F. Supp. 3d 553 (No. 5:15-CV-156-D). 109. RWCA I, 166 F. Supp. at 572. 110. Id. at 571–72. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

40 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3 at-large districts, thus allowing voters to vote for three school board members.111 Senate Bill 325 was wildly unpopular with Wake County residents and elected officials.112 In a public hearing hosted by the county’s legislative delegation on March 25, 2013, every member of the public who addressed the bill spoke against it.113 In April 2013, the school board adopted a resolution reaffirming its support for the current election process for its members and detailing why that process should be retained.114 Comments from the public during legislative committee hearings were also overwhelmingly in opposition to the bill.115 Grassroots groups like the Raleigh Wake Citizens Association(“RWCA”) and Concerned Citizens for African-American Children (“CCAAC”) organized resistance to the bill and attendance at all legislative hearings.116 Members of these groups and other community activists sought legal assistance in evaluating the new plan and devising a legal strategy for opposing it.

B. Filing Wright v. North Carolina Two months after Senate Bill 325 was ratified, the grassroots groups and individuals who had organized against the proposed legislation took action. They filed a lawsuit claiming that the law violated the equal protection guarantees of the state and federal constitutions because it created an election system that unjustifiably weighted the vote of some voters in the county much more heavily than others, thus violating the OPOV principle. Importantly, the groups marshalling the legal challenge have been intimately involved in advocating for the county’s socioeconomic diversity policy.117 For example, Concerned Citizens for African- American Children is a community-based organization dedicated to teaching parents about the policies, procedures, and laws that govern

111. Id. at 572–73. 112. Transcript of Testimony of Rev. Earl Johnson, supra note 91, at 25:22-28:4. 113. Transcript of Testimony of Sch. Bd. Member Christine Kushner, supra note 93, at 107:20-24. 114. Transcript of Testimony of Sch. Bd. Member Bill Fletcher, supra note 87, at 89:14-18. 115. See, e.g., Transcript of House Elections Committee Meeting, May 29, 2013 (on file with author) (statements of members of the public speaking out against the bill). 116. Transcript of Testimony of Rev. Earl Johnson, Dec. 16, 2015, supra note 91, at 25:22- 28:4. 117. Complaint at 6, Wright v. North Carolina, 975 F. Supp. 2d 539 (E.D.N.C. 2014) (No. 5:13-CV-607). RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 41 the children who attend the Wake County Public School System.118 As a parent-based organization, CCAAC works to ensure that all children receive equitable educational opportunities, which lead them to fight against racially segregated schools and school discipline policies that disparately impacted children of color.119 Similarly, the Raleigh Wake Citizens Association is Raleigh’s oldest African-American political organization and aims to empower minorities in Raleigh and Wake County.120 Since its beginning, the RWCA’s purpose has been to protect, encourage, educate, and help the citizens of Raleigh and Wake County in their civic, economic, social, educational and political advancement. RWCA members have been active in advocating for high quality education for all children.121 The theory of this suit was that Senate Bill 325 created a redistricting scheme similar to the one struck down in Larios.122 That is, while recognizing that the challenged plan had less than a ten percent overall deviation, some of the same patterns seen in Larios were also present here. In Larios, the Democratic plan favored urban and rural voters over suburban by underpopulating districts in the former areas—here, the plan favored suburban voters by placing them into underpopulated districts.123 Likewise, as in Larios, where Republican incumbents were targeted for defeat by being paired in a district with another incumbent, incumbents here who were registered Democrats or otherwise favored progressive policies were either paired or placed into politically-hostile districts.124 And, like in Larios, the plan was designed to secure electoral success for a party whose voters were not numerous enough to warrant extensive electoral success—that is, the plan illegitimately favored one political party.125 Thus, because the deviations in the Wake school board plan existed only to further illegitimate redistricting goals, this plan violated the one-person, one vote guarantee of the Fourteenth Amendment just like the Georgia state legislative redistricting plans did.126 Relying on

118. Id. at 5. 119. Id. at 6. 120. Id. 121. Id. 122. Id. at 2–3. 123. RWCA II, 827 F.3d 333, 351 (4th Cir. 2016); Larios v. Cox, 300 F. Supp. 2d 1320, 1327– 28 (N.D. Ga. 2004). 124. See Wright v. North Carolina, 787 F.3d 256, 265 (4th Cir. 2015); Larios, 300 F. Supp. 2d at 1329. 125. RWCA II, 827 F.3d at 346–47; Larios, 300 F. Supp. 2d at 1331. 126. Complaint at 19–20, Wright, 975 F. Supp. 2d 539 (No. 5:13-CV-607). RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

42 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3 guidance derived from Roman and then applied by the Fourth Circuit in Daly v. Hunt,127 the plaintiffs argued that plans with deviations under ten percent could be found invalid if the deviations derive from arbitrary, discriminatory or bad faith considerations.128 Applying a fairly standard Fourteenth Amendment intentional discrimination analysis, the challengers planned to demonstrate that any alleged justifications were prextextual or nonsensical, and thus it was the illegitimate motives, not the allegedly legitimate ones, that drove the deviations in the enacted plan.129 However, the district court hearing the Wright case rejected plaintiffs’ framing of the case. In early 2014, the district court dismissed plaintiffs’ complaint for failing to state a claim. It said, “[t]o claim impermissible political bias is to claim political gerrymandering. To claim political gerrymandering is to raise a claim that is nonjusticiable. Plaintiffs’ attempt to dress a political gerrymandering claim in OPOV clothing fails to state a claim for which this Court may grant relief.”130 The district court was both wrong and right, at least in a certain sense—the motivations being attacked were certainly partisan, but the constitutional injury was different.131 Unequally populated districts create a harm distinct from a redistricting plan that, as a whole, was intended to and has the effect of discriminating against voters of a particular party. Plaintiffs appealed that ruling to the Court of Appeals for the Fourth Circuit.

C. The 2015 Legislative Process Although the decision in Wright was still pending, during the early part of the 2015 North Carolina General Assembly legislative session, the legislature did not feel compelled to wait for a ruling from the circuit court before replicating the same strategies employed in the Wake County school board case. The Wake County Board of Commissioners quickly became the next target. While Republicans had enjoyed narrow control over the county commission when they urged changes to the Wake County school board, that control was ephemeral. Democrats in 2014 seized back control of the County Commission, and

127. 93 F.3d 1212 (4th Cir. 1996). 128. Wright, 787 F.3d at 265. 129. Complaint at 19–20, Wright, 975 F. Supp. 2d 539 (No. 5:13-CV-607). The Fourteenth Amendment intentional discrimination standard was established in Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 266 (1977). 130. Wright, 975 F. Supp. 2d at 546 (internal citation omitted). 131. Id. at 545–46. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 43 in 2015, all members of the County Commission were Democrats, and two were African-Americans.132 Unlike the school board, the seven members of the county commission were all elected at-large. And unlike the school board elections, county commission elections were partisan elections.133 The swing of control of the Board to Democrats did not go unnoticed by the Republican-controlled legislature. On March 4, 2015, Republican Senator Chad Barefoot of Wake and Franklin Counties introduced Senate Bill 181.134 The local bill imposed the same electoral structure and districts set to be used for the Wake County school board on the Board of County Commissioners.135 The districts were identical, and thus the bill would increase the County Commission by two seats and favor Republicans in five of the nine districts.136 The bill was ratified in April of 2015.137 Legislators used some of the same justifications for Senate Bill 181 that they used for Senate Bill 325, although there were some interesting differences.138 Again, legislators argued that the “doughnut” district would give suburban voters more voice in local government.139 However, proponents now raised the concern that it was too expensive for candidates to run countywide, and that moving to districts would alleviate that.140 Candidates running in the “doughnut” district would still essentially run countywide because the district spanned the entire perimeter of the county.141 Proponents also argued that moving from at-large elections to elections from districts only was necessary to avoid a lawsuit under the Voting Rights Act.142 Finally, proponents argued that it would be advantageous to have school board and county commission districts perfectly align, even though the appeal in Wright had already been argued and was pending decision.143 Just as with the school board redistricting bill, Wake County voters

132. RWCA I, 166 F. Supp. 3d at 574–75. 133. Id. at 573. 134. Id. at 575. 135. Id. 136. Id. at 573, 575; Transcript of Testimony of Dr. Jowei Chen, Dec. 17, 2015, at 45:21–46:2, RWCA I, 166 F. Supp. 3d 553 (No. 5:15-CV-156-D). 137. RWCA I, 166 F. Supp. 3d 553. 138. Id. at 575–79. 139. Id. at 575–76. 140. Id. at 578. 141. Id. at 569. 142. Id. at 578. 143. Id. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

44 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3 were outraged. Once again, RWCA, CCAC and other grassroots groups encouraged their members to attend legislative meetings on the bill and speak out against it.144 A local polling firm conducted a poll which indicated that the vast majority of Wake County voters were opposed to the change.145 Representative Rosa Gill led the legislative opposition to the bill.146 Gill spoke elegantly on the harms created by unnecessarily packing black voters into districts and by manipulating population deviations for partisan advantage.147 Just days after Senate Bill 181 was enacted, many of the same plaintiffs who brought the Wright case filed another lawsuit challenging Senate Bill 181.148

D. The Fourth Circuit’s Wright Decision One month after Senate Bill 181 was enacted, the Fourth Circuit reversed the motion to dismiss in Wright.149 The appeals court noted that Rule 12(b)(6) dismissals are disfavored where a complaint sets forth a novel legal theory, thus rejecting the district court’s ruling.150 The court relied on Roman for the proposition that being within ten percent overall deviation does not insulate a plan from attack on OPOV grounds, and found that the complaint sufficiently alleged that the plan was tainted by arbitrariness or discrimination.151 It explained: Plaintiffs allege in detail a redistricting that resulted in a maximum population deviation of nearly ten percent. Plaintiffs describe how and why that deviation was unjustified, discriminatory and unconstitutional. They do not allege that the apportionment plan with a maximum population deviation just barely under ten percent by itself supports their equal protection claim, but rather they plead facts indicating that the apportionment had a taint of arbitrariness or discrimination.152 Furthermore, the Fourth Circuit flatly rejected attempts by the district court and Appellees to distinguish Larios, which it found to be

144. See, e.g., Transcript, Senate Redistricting Committee Meeting, Mar. 10, 2015 (on file with author) (statements of members of the public speaking out against the bill). 145. Transcript of Testimony of Tom Jensen, Dec. 16, 2015, at 212:1-24, RWCA I, 166 F. Supp. 3d 553 (No. 5:15-CV-156-D). 146. See Transcript, House Floor Debate, Apr. 11, 2015, at 8–12 (on file with author) (debate by Rep. Rosa Gill). 147. Id. 148. See generally Complaint at 1–18, RWCA I, 166 F. Supp. 3d 553 (E.D.N.C. 2016) (No. 5:15-CV-156). 149. Wright v. North Carolina, 787 F.3d 256, 269–70 (4th Cir. 2015). 150. Id. at 263. 151. Id. at 264. 152. Id. at 265 (internal quotations omitted). RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 45

“notably” similar to Wright.153 Recognizing the limited precedential value of the summary affirmance, the Fourth Circuit nonetheless concluded that the Larios summary affirmance stood for the proposition that the ten percent rule did not create a safe harbor from OPOV challenge.154 The Fourth Circuit also noted that the Supreme Court agreed that discriminatory treatment of incumbents and favoring of regional areas would violate the Fourteenth Amendment if they explained the deviations in the challenged plan.155 Finally, the Fourth Circuit ruled that the district court’s dismissal of the case on the grounds that it was a nonjusticiable partisan gerrymander was wrong on two fronts. First, the district court misapprehended Vieth—because Vieth’s lead opinion was only a plurality opinion, partisan gerrymandering cases remain justiciable.156 Second, and more significantly, the appeals court explained that an OPOV case under the Equal Protection Clause was jurisprudentially distinct from a partisan gerrymandering case, and the Wright plaintiffs properly pled it as such.157

E. Filing of RWCA and Trial on the Consolidated Cases When filed on April 9, 2015, RWCA embraced largely the same theory as Wright.158 The overall deviation for the county commission plan was the same as it was for the school board plan, hovering just below that legally-significant ten percent threshold.159 However, in the county commission case, there was even more pretext for the justifications used to support that redistricting as compared to the school board redistricting.160 For example, legislative proponents argued that Senate Bill 325 created a superior system for electing school board members because it would allow voters to vote for more than one member, thus increasing the number of members who were responsive and accountable to voters.161 But that same logic did not seem to apply to the county commission. Since all seats were at-large, voters had seven members who were responsive and accountable to

153. Id. at 266. 154. Id. at 267. 155. Id. at 265 (citing Cox v. Larios, 542 U.S. 947, 949–50 (2004) (Stevens, J., concurring)). 156. Id. at 267. 157. Id. 158. Complaint at 2, RWCA I, 166 F. Supp. 3d 553 (E.D.N.C. 2016) (No. 5:15-CV-156). 159. RWCA I, 166 F. Supp. 3d at 575. 160. RWCA II, 827 F.3d 333, 349–50 (4th Cir. 2016) 161. RWCA I, 166 F. Supp. 3d at 572. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

46 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3 them, and Senate Bill 181 reduced the number of representatives from seven to two.162 Opponents of the bills of course noticed the inconsistency between the two bills during the legislative process. In addition to raising claims identical to the ones in Wright, the RWCA plaintiffs also alleged that the one majority black district created by the plan was, in the context of moving from at-large to district elections for the County Commission, an unconstitutional racial gerrymander.163 One of the seven single-member numbered districts was drawn to be majority black.164 Proponents said that such a district was necessary for the county commission because of the threat of a lawsuit and to ensure that black voters could elect their candidates of choice.165 Of course, that same majority black district existed in the school board redistricting plan, but school board members had long been elected from single-member districts, and that racial reason was not a purported justification for Senate Bill 325.166 The two cases were consolidated under RWCA in October 2015, and the court set an expedited schedule for discovery and trial because candidate filing for seats under the new county commission plan opened in December and closed in mid-January.167 The cases went to trial on December 16-18, 2015.168 Interestingly, at trial, the only defendant in appearance was the Wake County Board of Elections—an entity that was deemed the proper party because it was charged with implementing the legislatively-enacted plan, but had no involvement in designing or enacting the constitutionally-flawed plans.169 In Wright, the Fourth Circuit held that plaintiffs could not sue legislative leaders because they had no role in implementing the law.170 Neither legislative leaders nor the state legislators responsible for introduction of the challenged bills ever intervened. Thus, the Wake County Board of Elections was tasked

162. RWCA II, 827 F.3d at 349. 163. RWCA I, 166 F. Supp. 3d at 620. 164. Id. at 578. 165. Id. 166. See id. at 563, 572, 578 (noting that school board members were elected from single- member districts, and listing proponent justifications for the school board redistricting plan; noting that racial reasons were listed as a reason for moving to districts in the county commission). 167. Id. at 562. 168. Id. 169. See Wright v. North Carolina, 787 F. 3d 256, 262 (4th Cir. 2015) (“[T]he county Board of Elections . . . has the specific duty to enforce the challenged redistricting plan.”). 170. Id. at 261–62. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 47 with defending a law that it never requested or wanted.171 The Board of Elections had no direct insight into the reasons motivating the enacted bill. While it sought to entice state legislators to testify, those legislators claimed legislative privilege and declined to participate.172 The Board of Elections nevertheless vigorously cross examined plaintiffs’ witnesses and offered legal arguments that plaintiffs did not satisfy their legal burden in the OPOV case.173 At trial, plaintiffs called fifteen witnesses—two expert witnesses, Democratic state legislators, members of the school board and county commission, and several of the plaintiffs themselves. Importantly, plaintiffs’ critical expert witness, Dr. Jowei Chen of the University of Michigan, ran a large number of simulations for Wake County redistricting maps.174 In those simulations, Dr. Chen was able to hold constant several traditional redistricting criteria, such as compactness, equal population and respect for political subdivisions.175 He found that none of the hundreds of simulations achieved the partisan outcome achieved by the enacted plan.176 It was only by allowing the deviations in the simulated plans to rise to nearly ten percent that the legislature could achieve its skewed political product.177 Dr. Chen thus concluded that a desire to obtain partisan advantage motivated the deviations in the enacted plan.178 In February, the district court ruled against plaintiffs on all claims.179 In its decision, the court explicitly discounted all of plaintiffs’ witnesses as not credible, even those who were providing objective, factual evidence.180 The court ruled that “[i]n order to prove a prima facie case in a one person one vote challenge, plaintiffs must at least negate the most common legitimate reasons that could explain the legislature’s action.”181 Thus, the court applied rational basis review to the analysis, noting that “any conceivable legislative purpose is sufficient” to explain the redistricting plan and plaintiffs “have the burden to negate every

171. Transcript of Defense Opening Statement, Dec. 16, 2015, at 14, RWCA I, 166 F. Supp. 3d 553 (No. 5:15-CV-156-D). 172. Id. 173. Id. at 14–15. 174. Transcript of Testimony of Dr. Jowei Chen, supra note 136, at at 46:19–47:1. 175. Id. 176. Id. 177. Id. 178. Id. 179. RWCA I, 166 F. Supp. 3d 553, 560 (E.D.N.C. 2016). 180. See id. at 604–05. 181. Id. at 589. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

48 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3 conceivable basis which might support it.”182 The district court barely cited the Wright case, let alone followed its guidance on how a court should analyze OPOV claims. Plaintiffs promptly filed a notice of appeal.183

IV. REFINEMENT IN OPOV/PARTISAN GERRYMANDERING LAW DURING THE WAKE COUNTY LITIGATION As is often the case in the voting rights world, the applicable jurisprudence did not remain static during the pendency of litigation in the Wake County case. Just days before its trial commenced, the Supreme Court heard oral argument in Harris v. Arizona Independent Redistricting Commission,184 and the case was decided after the district court ruled, but before the Fourth Circuit would take the case up a second time. In Harris, voters challenged Arizona’s state legislative redistricting plans on OPOV grounds.185 The court unanimously affirmed the district court’s ruling upholding the plans.186 While plaintiffs alleged that the 4.07% deviation reflected the Commission’s attempt to favor the Democratic Party, the court below found that the deviations reflected the Commission’s efforts to comply with Section 5 of the Voting Rights Act.187 However, the Harris Court reaffirmed the conclusion that the Wake County litigants had drawn from the Larios summary affirmance. Accordingly, a plan with less than ten percent overall deviation was unconstitutional (and thus could be successfully challenged) if “it is more probable than not that a deviation of less than ten percent reflects the predominance of illegitimate reapportionment factors rather than the ‘legitimate considerations’ to which we have referred in Reynolds and later cases.”188 The Court went on to list those considerations it deemed legitimate: (1) “traditional districting principles such as compactness [and] contiguity,” (2) “maintaining the integrity of political subdivisions,” (3) complying with the Voting Rights Act,189 and (4) seeking “competitive balance among political parties.”190

182. Id. at 598. 183. Notice of Appeal, RWCA I, No. 5:15-CV-156-D (E.D.N.C. Mar. 14, 2016), ECF No. 66. 184. 136 S. Ct. 1301 (2016). 185. Id. at 1305. 186. Id. 187. Id. at 1307–08. 188. Id. at 1307. 189. Id. at 1306. 190. Id. (citing Gaffney v. Cummings, 412 U.S. 735, 752 (1973)). RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 49

With respect to that last criterion, the Court held in Gaffney v. Cummings191 that the advantage of one party over another is not “competivie balance among the political parties.”192 Gaffney involved redistricting after a decennial census, not mid-decade redistricting.193 The state legislative plans developed in Gaffney had an overall deviation of 1.81% in the Senate and 7.83% in the House.194 The redistricting Board responsible for redistricting in Gaffney explained that they followed a “policy of ‘political fairness,’ which aimed at a rough scheme of proportional representation of the two major political parties.”195 Importantly, in Gaffney there was no allegation or evidence presented that in order to achieve that goal, the Board systematically under- or over-populated districts controlled by one political party.196 For purposes of the analysis in Harris, the Court assumed, without deciding, that partisanship is an illegitimate criteria.197 Finally, the Court also noted that because of the challenge in weighing the dominance of legitimate redistricting criteria against illegitimate ones, “attacks on deviations under 10% will succeed only rarely, in unusual cases.”198 The state of the law in the partisan gerrymandering realm was moving too. After many years of struggling to develop articulable standards for courts presented with partisan gerrymandering cases, litigants, law professors, and social science experts finally started to coalesce around several possible ways of framing the legal problem, which manifested in actual cases.199 The timing is likely explained, too,

191. 412 U.S. 735 (1973). 192. Id. at 752. 193. Id. at 736. 194. See id. at 737. 195. Id. at 738. 196. See id. 197. See Harris v. Ariz. Indep. Redistricting Comm’n, 136 S. Ct. 1301, 1310 (2016) (noting that Section 5 was the law when Arizona redistricted in 2010). 198. Id. at 1307. 199. See, e.g., Jowei Chen & Jonathan Rodden, Cutting Through the Thicket: Redistricting Simulations and the Detection of Partisan Gerrymanders, 14 ELECTION L.J. 331, 331–45 (2015) (presenting a “redistricting algorithm [that] can be used to generate a benchmark against which to contrast a plan that has been called into constitutional question, thus laying bare any partisan advantage that cannot be attributed to legitimate legislative objectives”); Samuel S. H. Wang, Three Tests for Practical Evaluation of Partisan Gerrymandering, 68 STAN. L. REV., 1263, 1263 (2016) (proposing “three statistical tests to reliably assess asymmetry in state-level districting schemes: (1) an unrepresentative distortion in the number of seats won based on expectations from nationwide district characteristics; (2) a discrepancy in winning vote margins between the two parties; and (3) the construction of reliable wins for the party in charge of redistricting, as measured by either the difference between mean and median vote share, or an unusually even distribution of votes across districts”). RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

50 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3 in terms of increasing technology in the post-2010 redistricting cycle and an increasing belief on the part of map drawers that because no court had yet come up with a judicially-manageable standard for assessing partisan gerrymandering, there were no real checks on political gerrymanders.200 Such legislators could be more brazen, thinking that they would never be held accountable.201 Of course, these cases were not decided before RWCA, but the timing is nonetheless significant. Two cases filed before the RWCA decision showed significant potential for opening avenues of attack under a partisan gerrymandering theory—one in Wisconsin and one in Maryland. These cases articulate two very different theories of the case. While it may be unlikely that both turn out to be correct, there are at least two plausible standards percolating for later review by the Supreme Court. One approach, using a Fourteenth Amendment theory, is relevant to discussion here. That case, Whitford v. Gill,202 involved a challenge to Wisconsin’s 2011-enacted state legislative districts, and employs an equal protection-focused “efficiency gap” (“EG”) theory. The Republican- controlled legislature in Wisconsin redrew the state’s assembly districts, resulting in 2012 election results where the Republican Party received 48.6% of the two party statewide vote share for assembly candidates and won 60 of the 99 seats in the Wisconsin Assembly.203 Put most plainly, “[t]he efficiency gap is the difference between the parties’ respective wasted votes in an election, divided by the total number of votes cast.”204 That is, the proposed standard measures how efficient (or not) each voter’s vote is. Specifically, “it is the comparative relationship of one party’s wasted votes to another’s that yields the EG measure,” not the numbers standing alone, that highlight the constitutional flaw.205

200. See, e.g., Ruth Greenwood, Partisan Gerrymandering in 2016: More Extreme Than Ever Before, CAMPAIGN LEGAL CTR., http://www.campaignlegalcenter.org/sites/default/files/Post% 202016%20Election%20Analysis%20on%20Partisan%20Gerrymandering%20Impact.pdf (noting “redistricting plans in the post-2010 cycle evidence more extreme partisan gerrymandering than any other decade in modern American history”); Thomas E. Mann, We Must Address Gerrymandering, TIME (Oct. 13, 2016), http://time.com/collection- post/4527291/2016-election-gerrymandering/ (noting that litigators are working on providing the court with justiciable standards for assessing partisan gerrymanders in the light of the Court’s failure to identify one yet). 201. See generally sources cited supra note 200. 202. No. 15-cv-421, 2016 U.S. Dist. LEXIS 160811 (W.D. Wis. Nov. 21, 2016). 203. Id. at *29–30. 204. Id. at *31. 205. Id. at *168. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 51

As might be expected, defendants in Whitford filed a motion to dismiss, arguing that the efficiency gap was really just the “proportional-representation standard rejected by the Supreme Court in Vieth v. Jubelirer,” restated, and thus foreclosed.206 The court rejected this argument, explaining that it needed to hear from the experts and better understand the metric before it could make that conclusion.207 After trial, the court concluded that the metric was not just the proportional representation standard restated, and, more significantly, “[t]o say that the Constitution does not require proportional representation is not to say that highly disproportional representation may not be evidence of a discriminatory effect.”208 Ultimately, after hearing all the evidence, the court in Wisconsin concluded that with “plaintiffs’ proposed measure of asymmetry, the efficiency gap (or “EG”), the plaintiffs have “show[n] a burden, as measured by a reliable standard, on [their] representational rights.”209 Defendants only filed a notice of appeal on February 24, 2017, and thus it will be some time before the Supreme Court decides whether to hear the case.210

V. THE FOURTH CIRCUIT RULING IN RWCA The Fourth Circuit reversed the district court and instructed it to enter judgment for the plaintiffs on both their state and federal OPOV claims.211 With respect to the applicable legal standards, the Fourth Circuit explained that while Harris was decided after the district court ruling, the district court failed to avail itself of, and indeed ignored, the Fourth Circuit interpretation of the law as set forth in Wright.212 First, and perhaps most significantly, the Fourth Circuit recognized that the district court incorrectly applied “rational-basis review of whether a rational state policy could explain the redistricting generally,” rather than the “specific, deviation-focused inquiry” mandated by Wright and Harris.213 Plaintiffs do not have to negate any conceivable legislative purpose to support the challenged plan.214 Indeed, it is now clear that

206. Id. at *35. 207. See id. at *36. 208. Id. at *176. 209. Id. at *95 (citing League of United Latin Am. Citizens v. Perry, 548 U.S. 399, 418 (2006)). 210. Defendants’ Notice of Appeal, Whitford, ECF No. 191, No. 15-cv-421. 211. RWCA II, 827 F.3d 333, 338 (4th Cir. 2016). The Fourth Circuit held that the district court did not clearly err in rejecting the racial gerrymandering claim, and thus affirmed that portion of the lower court’s ruling. Id. 212. Id. at 341–42. 213. Id. at 342. 214. Id. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

52 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3 the Fourth Circuit articulated the proper legal standard in Wright, and Harris only confirmed that: plaintiffs in an OPOV case with less than 10% deviation must show by a preponderance of evidence that improper considerations predominate in explaining the deviations. If they do that, they have mounted a successful challenge. Second, the Fourth Circuit explained that the district court also erred legally by ignoring the relevance of Larios, which had been explained in Wright.215 Instead, the district court placed “heavy emphasis on Justice Scalia’s Larios dissent—an opinion with no precedential value.” Even worse, the “district court misapplied the core principles of Larios.”216 Contrary to the district court’s assertion, the Fourth Circuit explained that Larios does not require challengers to show disregard for “all districting principles,” and plaintiffs in the instant case should not have been required to demonstrate that every traditional redistricting criteria was disregarded across the plan.217 In short, the Fourth Circuit affirmed plaintiffs’ theory of the case and that Larios-style attacks on malapportionment employed for partisan gain does create equal protection violations. But the errors were not just legal. The Fourth Circuit incredulously noted that the district court “discounted every single one of Plaintiffs’ fifteen trial witnesses,” and that such discounting was clear error.218 The district court erred in discrediting entirely the testimony of legislators opposed to the challenged laws, even regarding objective facts.219 Most importantly, the district court clearly erred in rejecting Plaintiffs’ expert witness Dr. Jowei Chen.220 The Fourth Circuit explained that it was not that Dr. Chen’s simulations were simply “better plans” or that they were legally required, but that they allowed plaintiffs to demonstrate that illegitimate motivations, not legitimate criteria, caused the deviations.221 The appeals court recognized that plaintiffs showed, via Dr. Chen’s testimony, that the overall deviation on the plan was caused by an intent to create Republican advantage in the electoral districts.222 Moreover, following the logic of Roman and Daly, the Fourth Circuit

215. Id. 216. Id. 217. Id. at 343. 218. Id. 219. Id. 220. Id. at 344. 221. Id. 222. Id. at 344–45. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 53 noted that because so many of the proffered justifications were plainly pretextual, it was quite easy to conclude that illegitimate goals motivated the districts and the process as a whole.223 For example, in considering whether the school board redistricting created better alignment between electoral districts and student assignment districts, the appellate court actually looked at the maps in the record, and saw that the new electoral districts only exacerbated such mis-alignment.224 It thus concluded that such a justification was pretextual.225 The same result occurred when looking at each of the state’s justifications.226 Significantly, the Fourth Circuit reaffirmed that the appropriate standard not rational basis, but rather one that focused on the motivation for the overall deviations in the plan, and suggested, although not so directly, that consideration of the justifications proffered was relevant more for an understanding of whether the apportionment process as a whole was tainted by arbitrariness or discrimination.227 Finally, it noted that even had those justifications not been pretextual, none of them necessitated having high population deviations.228 Importantly, the Court of Appeals distinguished Gaffney and contrasted the legitimate consideration of “competitive balance among political parties” from what happened here.229 The court noted that the instant case, which involved a mid-decade redistricting as opposed to the normal post-census redistricting in Gaffney, was not an example of a legislature trying to ensure political fairness by creating a proportionate number of Republican and Democratic districts.230 That is what the map drawers in Gaffney were trying to do, but “the challenged redistricting here subverts political fairness and proportional representation and sublimates partisan gamesmanship.”231 The Court of Appeals thus concluded that challenged redistricting plans violated the Fourteenth Amendment, that the record supported only that conclusion, and ordered the district court to enter judgment for plaintiffs.232

223. See id. at 349. 224. See id. 225. Id. 226. See id. at 349–51. 227. See id. at 343. 228. See id. at 350–51. 229. See id. at 347, 355. 230. See id. 231. Id. at 347–48. 232. Id. at 345, 354. RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

54 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3

Defendant sought en banc review of the Fourth Circuit’s ruling, which was denied.233 Defendant did not seek certiorari in the United States Supreme Court, and thus the ruling stands in the Fourth Circuit.

VI. THE IMPLICATIONS OF RWCA ON PARTISAN GERRYMANDERING JURISPRUDENCE It might be easy to dismiss the significance of the case discussed here—these were only local bills being challenged, and plaintiffs did not face a defense with a vested interest in defending the challenged law. Indeed, such questioning may seem reasonable particularly given the Supreme Court’s dicta in Harris indicating that it would be the rare case where plaintiffs challenging redistricting plans with an overall deviation under ten percent succeed.234 However, when understanding how OPOV challenges operate in a landscape that includes more hyper-partisan interactions, and more sophisticated and technologically-advanced attempts to secure unfair partisan advantage, this case will have significant ripple effects not only in equipping litigants challenging malapportioned plans with powerful precedent but also possibly informing how partisan gerrymandering jurisprudence develops in the next few years. Indeed, in this hyper-partisan world, what the Court might think only rarely occurs (the sort of extreme manipulation of district lines seen in Larios and RWCA) is likely to become the new norm. Of course, dictum is not controlling law, and the Supreme Court may be convinced, if more similar cases are filed and litigated, that, in fact, such cases may be more than rarely successful if partisan-motivated malapportionment is a commonly-used tactic in redistricting. Thus, such dicta in Harris should not deter potential litigants. And it does not seem to have had that effect. The ruling has already been helpful to litigants in other cases. In Bethune-Hill v. Virginia State Board of Elections, a racially gerrymandering case that alleged that several Virginia House of Delegate districts were drawn predominantly on the basis of race, one defense asserted was that political considerations predominated in the drawing of the challenged districts.235 OneVirginia2021, a non-profit organization in Virginia

233. Order Denying Motion for Rehearing En Banc, RWCA II, 827 F.3d 333, 351 (4th Cir. 2016). 234. See Harris v. Ariz. Indep. Redistricting Comm’n, 136 S. Ct. 1301, 1307 (2016) (positing that “attacks on deviations under 10% will succeed only rarely, in unusual cases”). 235. 141 F. Supp. 3d 505, 542 (E.D. Va. 2015). RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

2017] THE ONLY CLEAR LIMITATION ON IMPROPER DISTRICTING 55 organized to promote a “comprehensive effort to remove gerrymandering from the redistricting process in Virginia,”236 submitted an amicus brief challenging that defense.237 Arguing that partisan gerrymandering should not be considered an acceptable defense to racial gerrymandering charges, the amicus pointed the Court to the RWCA case, urging the court that “in the face of evidence showing race-based redistricting, the state cannot defend a racial sorting by claiming that it was instead viewpoint discrimination.”238 Amicus emphasized that this case stood for the proposition that “partisan gerrymandering and viewpoint discrimination are not legitimate redistricting criteria.”239 This explicit connection between OPOV claims and partisan gerrymandering is significant. Most importantly, however, is the role that future aggressive OPOV litigation might have in directing the Court to developing a manageable standard for partisan gerrymandering. As suggested by the spate of cases now challenging redistricting plans as partisan gerrymanders,240 each employing new and more precise metrics for courts assessing those claims, many litigants are hopeful that the Supreme Court will finally be willing to establish the applicable standard and making such undemocratic efforts subject to judicial review. There is, of course, much uncertainty surrounding a potential new justice on the high court, and how the Court’s jurisprudence will change with the change in membership. However, it is important to remember that Justice Kennedy has been and likely will continue to be, for at least the near future, the critical vote in any partisan gerrymandering case. Justice Kennedy understands the problem with partisan gerrymanders as being very akin to the problem caused by plans that have unevenly populated districts—it is a question of fundamental fairness.241 As such, more litigation relying on the OPOV guarantee serves two purposes: (1) creating more precedent for the striking down of discriminatory plans with less than ten percent deviation; (2) fleshing

236. Brief for OneVirginia2021, Amicus Curiae Supporting Appellants at 1, Bethune-Hill v. Va. State Bd. of Elections, (No. 15-680) (Sep. 30, 2016). 237. Id. 238. Id. at 14. 239. Id. at 15. 240. See generally Whitford v. Gill, No. 15-cv-421, 2016 U.S. Dist. LEXIS 160811 (W.D. Wis. Nov. 21, 2016); League of Women Voters of N.C. v. Rucho, No. 1:16-CV-1164 (M.D.N.C. filed Sept. 22, 2016); Shapiro v. McManus, No. 1:13-cv-3233, 2016 U.S. Dist. LEXIS 112732 (D. Md. Aug. 24, 2016); Common Cause v. Rucho, No. 1:16-CV-1026 (M.D.N.C. Aug. 5, 2016). 241. See Vieth v. Jubelirer, 541 U.S. 267, 306–07 (Kennedy, J., concurring) (describing the need for a “fairness principle” in the instant case). RIGGS FINAL (DO NOT DELETE) 4/14/2017 7:41 PM

56 DUKE JOURNAL OF CONSTITUTIONAL LAW & PUBLIC POLICY [VOL. 12:3 out the types of expert and circumstantial evidence that can help a court understand how sophisticating partisan redistricting techniques can subvert political fairness. That is to say, when courts begin to understand how analyses like Dr. Chen’s RWCA analysis can provide strong circumstantial of partisan legislative intent, that creates a legal landscape in which partisan gerrymandering cases can flourish. These cases might also serve as a stepping stone for courts to become more comfortable with restricting improper partisan legislation, because courts can operate within the safety of the well-established OPOV framework, where the courts are not being asked to develop new standards. That type of precedent can also help litigants hone how they frame and articulate the harms wrought by political gerrymandering. And, ultimately, such litigation could very well inform Justice Kennedy’s decision, someday, to embrace a standard for measuring partisan gerrymanders. Litigants across the country must continue to rely on “the only clear limitation on proper districting practices,” and aggressively bring challenges to plans that use population disparities for improper partisan goals, in hopes that one day soon, the OPOV principle will no long be the “only” limitation on anti-democratic redistricting efforts.